ZCode with AI Gateway

ZCode is Z.ai's desktop agent development environment. Its custom providers speak the OpenAI-compatible Chat Completions format, which you can point at AI Gateway for unified model access and spend monitoring.

Quit ZCode, then run the Vercel CLI setup command:

npx vercel ai-gateway setup --agent zcode

The command writes a gateway provider and a seeded model list into ~/.zcode/v2/config.json with 0600 permissions. Re-run it to refresh the model list.

If you can't use the Vercel CLI, configure ZCode manually:

1. Create an API key

   Go to the AI Gateway section in the Vercel dashboard sidebar and click API keys to create a new API key.

2. Quit ZCode

   Quit the ZCode desktop app before editing its config. ZCode keeps its provider registry in memory and rewrites ~/.zcode/v2/config.json when it saves settings or exits, silently discarding edits made while it runs.

3. Add the gateway provider

   Add the provider to ~/.zcode/v2/config.json under provider, with the API format on Chat completions:

   ~/.zcode/v2/config.json

   {
     "provider": {
       "vercel-ai-gateway": {
         "name": "Vercel AI Gateway",
         "kind": "openai-compatible",
         "options": {
           "apiKey": "your-ai-gateway-api-key",
           "baseURL": "https://ai-gateway.vercel.sh/coding-agent/v1",
           "apiKeyRequired": true
         },
         "source": "custom",
         "models": {
           "openai/gpt-6-astra": {
             "limit": { "context": 1000000, "output": 128000 }
           }
         }
       }
     }
   }

   ZCode has no environment-variable lookup for custom providers, so the key is stored in this file. Create it with mode 0600.

4. Start ZCode and pick a model

   Open ZCode and pick a model under Vercel AI Gateway in the chat's model picker. Add more gateway models with Manage models > Add model.

   On a fresh install, choose Skip for now on ZCode's welcome screen. Its API-key dropdown only takes Z.ai/BigModel accounts, and the gateway provider lives in Settings > Model settings.

View your usage, spend, and request activity in the AI Gateway section in the Vercel dashboard sidebar. See the observability documentation for more details.

• Browse the models catalog for available models
• Configure spend limits on your API key